This article was downloaded by: [Pennsylvania State University]On: 15 March 2013, At: 05:15Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK
Transactions of the American Fisheries SocietyPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/utaf20
Otolith Age Validation and Growth Estimation fromOxytetracycline-Marked and Recaptured American ShadWilliam J. Duffy a b , Richard S. McBride a , Michael L. Hendricks c & Kenneth Oliveira ba National Oceanic and Atmospheric Administration, Northeast Fisheries Science Center, 166Water Street, Woods Hole, Massachusetts, 02543, USAb Department of Biology, University of Massachusetts Dartmouth, 285 Old Westport RoadNorth Dartmouth, Massachusetts, 02747, USAc Pennsylvania Fish and Boat Commission, Benner Spring Fish Research Station, 1735 ShilohRoad, State College, Pennsylvania, 16801, USAVersion of record first published: 31 Oct 2012.
To cite this article: William J. Duffy , Richard S. McBride , Michael L. Hendricks & Kenneth Oliveira (2012): Otolith AgeValidation and Growth Estimation from Oxytetracycline-Marked and Recaptured American Shad, Transactions of the AmericanFisheries Society, 141:6, 1664-1671
To link to this article: http://dx.doi.org/10.1080/00028487.2012.720631
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form toanyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses shouldbe independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly inconnection with or arising out of the use of this material.
Transactions of the American Fisheries Society 141:16641671, 2012C American Fisheries Society 2012ISSN: 0002-8487 print / 1548-8659 onlineDOI: 10.1080/00028487.2012.720631
Otolith Age Validation and Growth Estimation fromOxytetracycline-Marked and Recaptured American Shad
William J. Duffy*National Oceanic and Atmospheric Administration, Northeast Fisheries Science Center,166 Water Street, Woods Hole, Massachusetts 02543, USA; and Department of Biology,University of Massachusetts Dartmouth, 285 Old Westport Road North Dartmouth,Massachusetts 02747, USA
Richard S. McBrideNational Oceanic and Atmospheric Administration, Northeast Fisheries Science Center,166 Water Street, Woods Hole, Massachusetts 02543, USA
Michael L. HendricksPennsylvania Fish and Boat Commission, Benner Spring Fish Research Station, 1735 Shiloh Road,State College, Pennsylvania 16801, USA
Kenneth OliveiraDepartment of Biology, University of Massachusetts Dartmouth,285 Old Westport Road North Dartmouth, Massachusetts 02747, USA
AbstractThis study validated a whole-otolith aging method using known-age American shad Alosa sapidissima from the
Delaware River system. Although scale ages are commonly used in autecological and assessment studies of Americanshad, scale ages from the same fish could not be validated. New data reported here used known-aged otoliths andscales available from shad marked by oxytetracycline as larvae in a hatchery and recaptured as adults on or near theirspawning ground. A subset of whole otoliths were examined and annulidefined as a pair of translucent and opaquebandswere counted using males and females ranging in age from 3 to 9 years. The reading and interpretationof annuli by the more experienced reader were accurate with respect to the known age, whether measured as thepercent agreement (PA = 80%) or Changs coefficient of variation (CV = 3.11). The use of otoliths provided moreaccurate results than scales obtained from the same fish (PA = 46%; CV = 7.84). A second reader, who had noprevious experience with this species, had lower performance scores but also performed better with otoliths; thisdemonstrated the need for training and testing when using the whole-otolith aging method. Growth modeling usingages of known-age juveniles and adults confirmed dimorphic growth. Females grew larger than males (von BertalanffyL = 552 and 495-mm FL, respectively). The maximum age observed for females was only slightly older than males(9 versus 8 years). The superiority of the otolith-based age method makes it difficult to compare our results witholder, scale-based demographic studies, but it represents an improved method for generating ages for future stockassessments.
American shad Alosa sapidissima are an economically andculturally valuable herring (Clupeidae) native to the east coast ofNorth America (Munroe 2002). Dramatic declines have been
*Corresponding author: email@example.comReceived February 15, 2012; accepted July 31, 2012Published online October 31, 2012
documented for nearly two centuries, and age-based assess-ments have been used to evaluate certain stocks that range fromCanada to Florida (Limburg et al. 2003; ASMFC 2007). Scale
5 15 M
OTOLITH AGE VALIDATION AND GROWTH ESTIMATION 1665
aging is the only validated method (Cating 1953; Judy 1960),although this stands in contrast to many other species for whichotolith aging methods replaced scale aging methods decadesago (Summerfelt and Hall 1987). Scales of American shad are,in fact, difficult to read; their accuracy has been challenged formore than a century (LaPointe 1958; ASMFC 2007). Early scaleaging methods were repeatedly discredited (Borodin 1924; Leim1925) until Cating (1953) developed a scale aging method thatwas validated by Judy (1960) in the Connecticut River. Nonethe-less, a recent blind study using scales of known-age fish froma Pennsylvania river system failed to accurately age Americanshad, once again calling into question broad-scale use of Cat-ings scale-aging method (McBride et al. 2005). Most recently,Duffy et al. (2011) concluded that the application of Catingsmethod to age American shad scales is inappropriate across itsentire native range.
The choice of aging method is not merely a matter of conve-nience or tradition since using inaccurate age and growth datacan have negative effects on biomass estimates and manage-ment efforts (Beamish and McFarlane 1983; 1987). One of themost cited case studies on aging error occurred on the westerncoast of Canada with the sablefish Anoplopoma fimbria. Usingan unvalidated scale aging method, the commercial fishery wasthought to comprise age-3 to age-8 fish; once a validated otolithmethod was developed, it was found that the commercial fisheryconsisted of fish ages 4 to 40 (Beamish and McFarlane 1987).Such aging errors lead to inaccurate estimates of mortality thatresult in annual catch recommendations too high for sustainabil-ity or too low for industry efficiency (Beamish and McFarlane1983; 1987).
Validation studies have increased over the last two decadessince Beamish and McFarlanes (1983) classic The ForgottenRequirement for Age Validation in Fisheries Biology was pub-lished. Since then, more emphasis has been put on validatingaging methods. There are many types of validation studies, butvery few of them validate all age-classes of a fish species, whichis the primary goal of any aging study (Beamish and McFarlane1983; Campana 2001). Most studies use marginal incrementanalysis, but for only one or a few of the youngest age-classes;however, validating the formation of one scale annulus per yearin this manner is usually insufficient for even fish with mediumlongevity because scale annuli are difficult to discern after a fewyears of growth (Beamish and McFarlane 1987). Other studiesuse tagging and recapturing of wild-caught fish to validate theformation of the annulus. This is accomplished by either usingan external tag or internally injecting the fish with a chem-ical marker (Beamish and McFarlane 1983; Campana 2001;DeCicco and Brown 2006). This method can confirm the fre-quency of annulus deposition, but it does not validate the abso-lute age of the fish because the age of the tagged fish is rarelyknown (Beamish and McFarlane 1983; Campana 2001).
In this study, we used otoliths of known-aged American shadcollected from the Delaware Bay ecosystem as an alternate agingstructure to scales. Aging American shad using otoliths has been
attempted, but the method has not been validated (Limburg2001; Aschenbach et al. 2006). We validate an otolith-basedaging method across all typical ages found on the spawninggrounds and simultaneously compare the results with a scale-based aging method. Finally, given the uncertainty of the agedata obtained from scales used in the previous stock assessmentfor Delaware River American shad (ASMFC 2007), we fit sizeand age of known-aged fish to the von Bertalanffy growth modelto update growth parameter estimates.
METHODSSample collection.We used a collection of known-
age American shad females and males available from thePennsylvania Fish and Boat Commission (PFBC). The PFBChas cultured and marked shad as larvae using oxytetracycline(OTC) since the mid-1980s (Hendricks et al. 1991). More than1,100 marked, known-age juveniles and adults were recapturedand sampled between 1995 and 2007 from the Delaware Riverecosystem, primarily from two of its tributaries: the LehighRiver and Schuylkill River.
Hatchery-raised shad larvae were repeatedly immersed inOTC to produce multiple marks on their otoliths (Hendrickset al. 1991). Most annual cohorts were given a unique seriesof OTC marks to identify the river of origin and cohort. Thesemarked fish were released as larvae into the Delaware Riverecosystem, where they spend a few months in freshwater beforemigrating out to sea as part of their anadromous life history.When juveniles (age 0) or adults were recaptured, fish size wasmeasured as fork length to the nearest millimeter. Also, foradults, 1020 scales were removed from the area just under thedorsal fin, and both sagittal otoliths were excised from the head.
Aging.One otolith was used to determine if the shad waswild or hatchery raised based on OTC markings, and the otherwas used to examine for annuli (Hendricks et al. 1991). Mostotoliths had been encased in epoxy resin, and the majority ofthese were difficult to read because of discoloration and crack-ing of the epoxy. A few dozen otoliths were stored in vialsand submerged in mineral oil, and were very suitable for aging.All otoliths stored in vials were used for the validation (if notbroken). Only some of those encased in epoxy could be used,specifically when the epoxy did not restrict visibility or distortthe appearance of the otolith (Table 1). Otoliths were not sec-tioned, sanded, or polished, so we refer to this as a whole-otolithaging method.
Matching scale samples were prepared when the otolith forthat fish was used, unless the scales were missing or regener-ated. Scales were prepared and scale annuli were recognizedfollowing the methods outlined by Duffy et al. (2011).
Shad otoliths were viewed under a dissecting microscopebetween 15 and 20 using reflected light. The otoliths wereaged by counting alternating translucent and opaque bandsfrom the otolith core to the margin of the postrostrum. Winter(translucent) bands appeared dark, and summer (opaque) bands
5 15 M
1666 DUFFY ET AL.
TABLE 1. Sample size, by age, for both the validation and the growth mod-eling aspects of this study. Otoliths and scales were from known-age fish andwere aged by both readers. The numbers in parentheses are otoliths encased inepoxy. The ages used in the growth model were calculated by time differencefrom hatching to recapture.
Otoliths Scales Growth modelAge N N N
0 0 0 6301 0 0 02 0 0 03 2 (2) 2 184 11 (10) 8 1105 35 (9) 25 2006 29 (12) 24 1347 9 (6) 6 228 1 (1) 1 89 1 (0) 1 1Total 88 (40) 67 1,123
appeared white (Penttila et al. 1988). Although we assumed thatthe band pairs were formed annually and we tested this as partof this study, the appearance of these band pairs changed fromthe core to the edge of the otolith. The first annulus boundeda broad translucent band extending from an opaque band justafter the nucleus of the otolith, which was dark under reflectedlight (Figure 1a). Between the first and second annulus, a diffusetranslucent band was often evident, but this did not represent afull year of growth (i.e., a false annulus; Figure 1b). Summerfeltand Hall (1987) define a false annulus as an abrupt discontinuityin a band with irregular spacing. The distance between the firstand second annulus also tended to be greater than the distancesbetween subsequent annuli. After the third year, the spacing be-tween translucent zones was fairly consistent in distance and theannuli were more coherent (Figure 1b). The edge of the otolithwas considered the final annulus because shad were caught ontheir spawning grounds.
Age validation.Two readers conducted blind studies of ac-curacy and precision using the otoliths and scales from the samefish. As in Summerfelt and Hall (1987), we define accuracy asthe proximity of an age estimate to the actual value, and preci-sion as the measure of consistency in age estimates when count-ing annuli more than once. The simplest performance measurewas to calculate percent agreement (PA) as the percentage ofages that agreed with either the known-age fish (i.e., accuracy)or within or between readers (i.e., precision), expressed as
PA = 100 AN
where A is the number of correct replicate ages (i.e., to the exactyear), and N is the total number of fish aged. A PA greater than80% was considered acceptable (Campana 2001).
Changs coefficient of variation (CV; Chang 1982) was alsoused as a second performance measure of precision and accu-racy. Changs CV first calculates the standard deviation of multi-ple age readings from a single fish, divides this by the mean ageof that fish, and, in the form we use, sums these values amongall fish aged as
CV = 100 1N
where N is the total number of fish aged, R is the number oftimes each fish is aged, Xij is the ith age determination of thejth fish, and Xj is the mean age estimate of the jth fish (Campanaet al. 1995). A CV < 5 was considered acceptable (Campana2001).
When PA was below 80% or Changs CV was greater than5, Bowkers (1948) test of symmetry was used to determine ifdeviations from the diagonal on an age frequency table (firstread versus second read) show syst...